- Email: [email protected]
Preparation of γ-Al2O3-supported molybdenum oxycarbide catalyst and its HDS activity
JOURNAL OF FUEL CHEMISTRY AND TECHNOLOGY Volume 36, Issue 3, June 2008 Online English edition of the Chinese language journal Cite this article as: J Fuel Chem Technol, 2008, 36(3), 316−321
RESEARCH PAPER
Preparation of γ-Al2O3-supported molybdenum oxycarbide catalyst and its HDS activity ZHU Quan-li1, *, ZHANG Qiu-lin2, KE Dong-xian1, YANG Xiao-bin1, CHEN Qi-feng2 1
Department of Chemistry, Hanshan Normal University, Chaozhou 521041, China
2
Chaozhou Xianglu Tungsten Industry Co., Ltd, Chaozhou 521000, China
Abstract: Investigation into the temperature programmed reduction (TPR) of bulk or γ-Al2 O3 supported molybdenum trioxide using n-hexane as a carbon source in hydrogen was carried out. T he resulting catalyst was evaluated using dibenzothiophene (DBT) as a model compound for the hydrodesulfurization (HDS). T he results show that the utilization of n-hexane can lower the temperature, to fulfill carburization to some extent. For DBT HDS, however, there exist s an optimal temperature at which carburization is carried out, leading to a high activity of the resulting catalyst. In these experiments, it has been found that the resulting catalyst exhibits the highest DBT HDS activity if the precursor, MoO3 /γ-Al2 O3, with 30% theoretic loading, is carburized at 620º C under atmospheric pressure and 0.025 mol ratio of n-hexane to hydrogen in carburizing reagents.
Key Words:
molybdenum oxycarbide; hydrodesulfurization; n-hexane; XRD
The molybdenum-based catalyst has been widely used for HDS of petroleum distillates and oil derived from coal in the industry. Thus far, its activity has been increased to a very high level via modifications in the past decades [1]. Early transition metal carbide, particularly of molybdenum and tungsten, has drawn much attention as it was reported to have catalytic activity similar to noble metal[2]. In fact, transition metal carbides have shown excellent catalytic activity in many chemical reactions [3–9]. It is even believed to be a commercial HDS catalyst of the future because of its excellent performance in HDS[10]. Molybdenum sulfide, carbide or oxycarbide catalysts can all be considered to be the result of modifications on a molybdenum-based catalyst. It is clear that different modifications, even every step of the modification, will lead to a resulting catalyst with a different surface structure, and furthermore with a different catalytic activity. Most of the carbide catalysts reported in the references have been prepared via the TPR method developed by Boudart et al[11], mostly using 20% of methane in hydrogen as the carburizing reagent. However, there are many hydrocarbons that can be used as the carbon source in carburization. The tests of C2, C3, and C4 hydrocarbons have been reported in references [12–14]. It can be inferred from these reports that the utilization of a longer chain hydrocarbon can lower the temperature, to fulfill the carburization to some extent.
Methane is the most widespread and most cheap hydrocarbon all over the world, but higher temperature is needed to transform metal oxide into carbide. C2, C3, and C4 hydrocarbons are of higher activity for carburization. However, they are put into commercial use very carefully because of the difficulty in acquiring and dealing with them. Therefore, the investigation into the possibility of liquid hydrocarbon being used as a carbon source is meaningful. In this article, the test of n-hexane in carburization of molybdenum and the activity of the resulting catalyst for DBT HDS are reported and discussed.
1 1.1
Experimental Catalysts preparation
Oxide precursors were prepared by impregnating γ-Al2O3 (20–40 mesh, BET specific area, 304 m2/g, a carrier of commercial HDS catalyst, Fushun Research Institute of Petroleum and Petrochemicals, SINOPEC) with an aqueous solution of a calculated quantity of ammonium heptamolybdate (A.R.) for 4 h. Following this step, the resulting powders were dried at 110ºC for 12 h. It was finally calcined in air at 500ºC for 3 h. Thus the oxide precursor MoO3/ γ-Al2O3 with 30% theoretical loading was obtained. Pure molybdenum trioxide was obtained by the calcination
Received: 23-Nov-2007; Revised: 28-Fan-2008 * Corresponding author. [email protected] Copyright 2008, Institute of Coal Chemistry, Chinese Academy of Sciences. Published by Elsevier Limited. All rights reserved.
ZHU Quan-li et al. / Journal of Fuel Chemistry and Technology, 2008, 36(3): 316−321
of ammonium heptamolybdate in air at 500ºC for 3 h. The apparatus used for the carburization and activity measurement was assembled by the authors. The system pressure can be controlled by a pressure regulator coupled with a back pressure regulator (TESCOM). The gas flow rate was regulated with a mass flow controller (Beijing Sevenstar Electronics Co., Ltd, P.R. China). The feeding rate of liquid materials was controlled by a liquid pump (Syltech 500). A stainless steel tubular reactor (8 mm, id) was adopted. The volatile components in the off-gas were monitored online by the Qic 20 gas analysis system (Hiden Analytical Ltd). Usually, 2.1 g of sample (ca. 3 mL) was loaded in the central part of the reactor with quartz particles (10–20 mesh) filling the spare parts. TPR of molybdenum oxide, supported or unsupported, was carried out under atmospheric pressure at a heating rate of 1ºC/min from 100ºC, to a final temperature, and then maintained at this temperature for a period of time. The ratio of n-hexane to hydrogen was achieved via changing the feeding rate of n-hexane, whereas, the hydrogen flow rate was fixed at 0.25 mol/h. In the TPR test, a Qic 20 gas analysis system was used to monitor the components in the effluent gas. Before the TPR test, the oxide precursor was dried at 115ºC for more than 12 h, and nitrogen sweeping was exempted. When the TPR was completed, the resulting catalyst was evaluated for DBT HDS directly, or it was cooled down to room temperature in hydrogen, which was followed by passivation with 1% O2 in N2 for 6 h. When DBT HDS was carried out at 290ºC under 3.0 MPa of hydrogen partial pressure, the solution of 1 mass % DBT in cyclohexane was fed at the rate of 9.4 mL/h, and the H2 flow rate was maintained at 0.12 mol/h. The liquid sample gathered at intervals of 2 h was analyzed using the ultraviolet fluorescence method (ZDS-2000, Jiangyan High-tech Analytical Instrument Co., Ltd, China) excepting the oil product gathered at the first hour. The liquid–liquid equilibrium can arrive at 97.5% or above, thus the hydrodesulfurization degree, denoted as HDS%, was simply calculated according to the following formula: HDS% =[1 – Sconcentration in sample (g-S/L)/S concentration in feedstock (g-S/L)]×100%
by X-ray diffraction using Philips X’Pert PRO diffractometer with CuKa radiation.
2 2.1
Results and discussion XRD Results
XRD patterns of molybdenum oxide precursors reduced at atmospheric pressure to different final temperatures under the condition mentioned earlier are shown in Fig. 1. It can be found in Fig. 1 that peaks at 26.0, 37.0, and 53.5o for sample a, were attributed to the characteristic diffraction peaks of MoO2 (JCPDS 86-0135). This means that MoO3 can only be reduced to MoO2. As shown in Fig. 1, the diffraction peaks at 39.5, 37.9, 34.5, 52.2, 61.6, and 69.6o for sample b, were attributed to β-Mo 2C[15], apart from the concomitant characteristic diffraction peaks of molybdenum dioxide. The XRD pattern of sample c was similar to that of sample b, but the intensity of the characteristic diffraction peaks of molybdenum dioxide was much weaker. In view of the passivation process leading to oxidation, to some extent, after carburization, it can be deemed that the carburization at 650ºC was almost complete. However, the case of the supported catalyst was different. As shown in Fig. 1 d, no characteristic diffraction peaks of molybdenum species were visible, except those of γ-alumina.
Fig. 1 XRD patterns of oxide precursors reduced at different temperatures a: ended at 550º C; b: ended at 600º C; c: ended at 650º C; d: ended at 650º C
1.2 Carburization samples preparation 2.2 Sample a, b, and c resulted from TPR of molybdenum trioxide and sample d from 30% MoO3/γ-Al2O3, using hydrogen mixed with n-hexane as the carburizing reagents at atmospheric pressure, and at a heating rate of 1 K/min from 373 K to a final temperature and held at this temperature for 2 h. When conducted, the molar ratio of n-hexane to hydrogen was kept at 0.025. The crystalline components of the catalyst were identified
TPR-MS of MoO 3 /γ-Al 2 O 3
The blank experiment was performed to detect the catalysis of the stainless steel reactor together with the quartz particles toward the decomposition of n-hexane. Fig. 2 shows that the ionic fragments resulted from volatile products tracked by the Qic 20 gas analysis system versus time on stream during temperature-programmed reduction, sweeping the reactor and quartz particles using the mixture of n-hexane
ZHU Quan-li et al. / Journal of Fuel Chemistry and Technology, 2008, 36(3): 316−321
and hydrogen. These ionic fragments shown in Fig. 2 can generally result from many molecules, but they, particularly the change in their intensity, can be limited within much lesser species if feasible reactions are taken into account. The profile of signal m/z = 78, attributed to benzene in the products, indicated that the stainless steel reactor cannot catalyze the dehydrocyclization of n-hexane to benzene below 650ºC. The profile of signal m/z = 18, attributed to water steam, indicates that some water can be removed from the quartz particles, mainly below approximately 430ºC. It is hardly formed above this temperature. Up to 600ºC, there is no change in m/z = 44 attributed to CO2 or C3 or higher hydrocarbon, which meant no reduction of quartz. A slight decrease from above 600°C can result in the decrease of C3 or higher hydrocarbon. The signal of m/z = 43 can result from C3 or higher hydrocarbon. Its decrease from above approximately 580ºC denoted n-hexane consumption rather than the substantial formation of C3 and C4 hydrocarbons; otherwise it increased with elevated temperature . The increase of m/z = 16, 28, and 30 above 600ºC, can occur in all organic compounds, which indicates the initiation of n-hexane decomposition to methane or C2 hydrocarbon, if not, there must be some heavier ionic fragments that increase with temperature. Fig. 2 shows the fact that n-hexane decomposition takes place above 600ºC in this condition, no matter how this stainless steel reactor including quartz filling catalyzes.
at its maximum at approximately 525ºC. As for the signal m/z = 44, though weak, it arrives at its highest at 550ºC. The signal of m/z = 43 shows a slight decrease in temperature before 250ºC. Following this temperature a relatively stable status is maintained up to approximately 460ºC. It decreases substantially with temperature above 460ºC.
Fig. 3 TPR – MS of MoO3 /γ -Al2 O3 The experiment was carried out under atmospheric pressure, using a carburizing reagent with 0.025 molar ratio of n-hexane to hydrogen, heating at a rate of 1ºC /min from 100ºC to 650ºC
2.3 Effect of compositions in carburizing reagents on DBT HDS activity The effect of hydrocarbon-to-hydrogen ratio in carburizing reagents on the DBT HDS activity of the resulting catalyst is shown in Fig. 4.
Fig. 2 Blank experiment The experiment was carried out at atmospheric pressure, using a carburizing reagent with 0.025 molar ratio of n-hexane to hydrogen, heating at a rate of 1ºC /min from 100ºC to 650ºC
The online analysis results of the volatile products during TPR of MoO3/γ-Al2O3 using the mixture of n-hexane and hydrogen are shown in Fig. 3. The profile of m/z=18 shows water formation with temperature. Three reduction peaks can be found at 340, 504, and 620ºC, respectively, and a should er peak at approximately 366 ºC. A peak at approximately 560 ºC occurs in the curve of m/z = 78, across the experimental temperature range. The signal m/z = 28 arrives at its highest at approximately 545ºC, whereas, the signal m/z = 30 arrives
Fig. 4 The effect of carburizing agent compositions on DBT HDS activity Catalysts that resulted from 30% MoO3 /γ -Al2 O3 carburized at atmospheric pressure using n-hexane as the carbon source, heating at a rate of 1º C /min from 100º C to 650ºC and held for 2 h at this final temperature. The feeding rate of n-hexane is shown in the legend
As mentioned earlier, the ratio of hydrocarbon to hydrogen in the carburizing reagents varied through changing the n-hexane feeding rate at a fixed hydrogen flow rate. DBT HDS activity of the resulting catalysts with time on stream in Fig. 4 showed a very high initial activity, even as high as
ZHU Quan-li et al. / Journal of Fuel Chemistry and Technology, 2008, 36(3): 316−321
97%, for every catalyst. However, these catalysts deactivated rapidly to approximately 47% after a 17-h run. However, the catalyst that resulted from carburization under the condition of 0.80 mL/h of n-hexane feeding rate and 0.025 molar ratio of n-hexane to hydrogen, guaranteed to be with higher activity within a relatively steady stage, judging from the activity trend shown in Fig. 4. For the catalyst resulting from the lowest n-hexane feeding rate, it was possible to get to a relatively steady stage of lowest activity. 2.4 Effect of carburization temperature and time on DBT HDS activity The optimal carburization temperature was about 620ºC, deduced from the results shown in Fig. 3, which will be discussed later. Therefore, several carburization temperatures scattered around 620ºC were investigated. The result is presented in Fig. 5. All catalysts showed very initial activity together with fast deactivation, as shown in Fig. 4. Judging by the activity trend, the catalyst that resulted from the carburization at 620ºC for 2 h showed a better activity than those carburized at 650ºC and 600ºC for 2 h, respectively, that is , promised to get to a steady state of a higher activity. When the final temperature ended at 620ºC, the catalyst resulted from 1 h of carburization, and exhibited a higher activity than those of 2 h and 4 h carburization, respectively.
Fig. 5 Effect of carburization temperature and time on HDS activity Catalysts that resulted from 30% MoO3 /γ -Al2 O3 carburized at atmospheric pressure using n-hexane as a carbon source with 0.025 molar ratio of n-hexane to hydrogen, heating at a rate of 1º C /min from 100º C to a different final temperature and held for different time shown in the legend at final temperature
2.5
Discussions
The pyrolysis or hydrocracking of n-hexane can lead to smaller hydrocarbon molecules, therefore, it is possible to use n-hexane as the carbon source, similar to methane, C2, C3, and C4 hydrocarbons, for the carburization of transition metal precursors . Comparing the ionic fragments resulting from
n-hexane, shown in Fig. 2 and Fig. 3, it can be found that γ-alumina supported molybdenum oxide catalyzed the n-hexane decomposition at a suitably high temperature, whether the stainless steel reactor had catalysis or not. Certainly, some transformation occurred on the surface of the molybdenum oxide with elevated temperature. In fact, the carburization of the molybdenum oxide was also a process of oxygen removal, thus the formation of an oxygen containing species could indicate the reduction state of molybdenum oxide. The profile of water formation in Fig. 3, m/z = 18, was similar to that reduced by hydrogen at a heating rate of 1 K/min in ref[16], and also similar to that carburized using butane and hydrogen [14]. Taking the XRD results shown in Fig. 1 into account, three reduction peaks below 550ºC were attributed to the process of MoO3 reduced to MoO2, whereas, oxygen removal above 550ºC was attributed to carburization. In Fig. 3, the signal of m/z = 44 could be attributed to concomitant ionic fragments of C3 or heavier hydrocarbon, as also to CO2. If it was the former, its intensity would synchronously vary with fragments such as , m/z = 43 and so on. However, the fact shown in Fig. 3 was not the case. This meant CO2 formation contributed to the increased intensity of m/z = 44. The signal of m/z = 28 could be attributed to C2 or heavier hydrocarbon, or N2 or CO. If N2 existed as an impurity, it did not change the intensity of m/z = 28. If the signals of m/z = 28 and 30 were only attributed to C2 hydrocarbon, they should arrive at the maximum at the same temperature. The results in Fig. 3 indicated that there were other species that contributed to m/z = 28 or 30. When CO2 was formed, gun with the CO formation. Prior to COx formation, the pyrolytic C or CHx group that resulted from the rupture intramolecule of n-hexane was deposited on the surface of molybdenum oxide. This led to the oxycarbide formation, namely, initiating the carburization. The signal of m/z = 78 in Fig. 3 indicated benzene formation at 560ºC, which resulted from the dehydrocyclization of n-hexane via cyclohexane intermediate over the molybdenum oxide. Accordingly, the surface of MoO3/γ-Alumina had been subjected to a process of MoO3 → MoO2, and an initiated carburization, as mentioned earlier. This dehydrocyclization of n-hexane occurred more possibly via bifunctional catalysis than by oxidative dehydrocyclization, catalyzed by a metal oxide[17], because the oxycarbide should be formed before the carbide formation[15]. Moreover, the oxycarbide rather than carbide possesses the bifunction[18]. When the oxycarbide was formed, the catalysis to hydrocracking was enhanced, and the enhanced catalysis resulted in rapid methane formation as shown in Fig. 3. The promoted deep hydrocracking in its turn enhanced the carburization. Therefore, carburization should be completed at a temperature above 560ºC. As shown in Fig. 1, the carburization was almost complete at 650ºC for bulk MoO3.
ZHU Quan-li et al. / Journal of Fuel Chemistry and Technology, 2008, 36(3): 316−321
In view of the promotion of support to the dispersion of MoO3, resulting in an easier chemical reaction, the carburization of MoO3/γ-Al2O3 would be favored. This was to say that MoO3/γ-Al2O3 could be highly carburized before 650ºC. Taking all factors into account, the temperature of 620ºC indicated by the profile of m/z = 18 signal in Fig. 3 was possibly the optimal carburization temperature for 30% MoO3/γ-Al2O3. Compared with methane, a lower temperature was needed to fulfill the carburization of the supported molybdenum oxide[19]. No matter what hydrocarbon is used as the carbon source, there is always an optimal ratio of hydrocarbon to hydrogen in the carburizing reagents for the resulting catalyst. As shown in Fig. 4, this optimal molar ratio was 0.025 for the carburization of 30% MoO3/γ-Al2O3 at 650ºC, when n-hexane was used as the carbon source, and the resulting catalyst was evaluated by DBT HDS. In view of the theoretically possible carbon deposition quantity, this value was less than 20% methane or 5% butane in hydrogen. It accounted for the higher chemical activity of n-hexane for the carburization. Of course, this value should be related to the oxide precursor, to the carburization temperature, and to the reaction the resulting catalyst should be oriented to, and so on. A higher n-hexane to hydrogen ratio in this work than reported earlier[20] was possibly attributed to the hydrogen used in the carburizing reagents. There are two parallel routes to desulfurization: one is direct desulfurization and the other is the desulfurization though hydrogenation [21]. Therefore, the surface acidity and hydrogenation capacity are needed for a good HDS catalyst. Usually, the carbide based catalyst reported in literature was a real oxycarbide catalyst because very high temperature was needed to complete the oxygen removal[19]. The difference among them was the carburization degree. The residual oxygen not only helped to stabilize the so-called carbide phase, but also helped to form B-acidic sites, to activate hydrogen or to promote the H spillover, which were helpful to HDS. The Fig. 5 indicated that the catalyst resulted from carburization at 620ºC rather than at 650ºC or at 600ºC had a better DBT HDS activity, and the carburization at 620ºC for 1 h was better than for 2 or 4 h. This result meant that some residual oxygen in the carbide catalyst was necessary to maintain a highly stable HDS activity. As reported in references [3,14] , a moderately carburized catalyst exhibited better hydrotreating activity. The higher temperature carburization resulted in the formation of a typical carbide of catalysis more similar to a noble metal, but other effects upon the activity, such as the decrease of specific area, surface acidic sites , and so on, worked at the same time. The results shown in Fig. 5 indicated that the optimal carburization temperature and time and the composition of carburizing reagents will result in an optimal synergism
among the specific area, pore distribution, surface acidity, hydrogenation ability, and so on. In fact, the molybdenum carbide catalyst in these experiments, whether carburized at 650ºC, at 600ºC or at another temperature, same as in most of the literature, was actually made up of oxycarbide. When n-hexane was used to carburize the metal oxide, a lower carburization temperature was needed because of its high chemical activity. However, it was this kind of high chemical activity that also led to substantial carbon deposition on the resulting catalyst at the same time. Thus the composition of carburizing reagents was needed to be more strictly controlled. 3
Conclusions
Primary study on molybdenum oxide carburization using n-hexane as carbon source was carried out. It was found that n-hexane could be used in the preparation of a carbide catalyst and a lower temperature was needed to fulfill the carburization. To obtain this kind of oxycarbide-based catalyst with high HDS activity, many parameters needed to be strictly controlled. For the precursor of 30% MoO3/γ-Al2O3, high DBT HDS activity was obtained if it was carburized at 620ºC for 1 h under atmospheric pressure, using carburizing reagents with 0.025 molar ratio of n-hexane to hydrogen. Certainly, the effects of many other preparation parameters on the catalytic perfomance are needed to be studied in detail.
Acknowledgment:
Zhu Quanli and his co-authors thank the
Chaozhou Xianglu Tungsten Industry Co., Ltd for the financial support rendered by it for this study.
References [1] Kabe T , Ishihara A, Qian W. Hydr odesulfurization and hydrodenitrogenation: Chemistry and Engineering. Weinheim: Wiley-VCH, 1999. [2] Levy R B, Boudart M. Platinum-like behavior of tungsten carbide in surface catalysis. Science, 1973, 181(4099): 547–549. [3] Jin G Z, Zhu J H, Fan X U, Sun G D, Gao J B. Effect of Ni promoter
on
dibenzothiophene
hydrodesulfurization
performance of molybdenum carbide catalyst . Chinese Journal of Catalysis, 2006, 27(10): 899–903. [4] Li S, Lee
J
S,
Hyeon
T,
Suslick
K
S.
Catalytic
hydrodenitrogenation of indole over molybdenum nitride and carbides with different structures. Appl Catal A, 1999, 184(1): 1–9. [5] Claridge J B, York A P E, Brungs A J, Marquez-alvarez C, Sloan J, Tsang S C, Green M L H. New catalysts for the conversion of methane to synthesis gas: Molybdenum and
ZHU Quan-li et al. / Journal of Fuel Chemistry and Technology, 2008, 36(3): 316−321 tungsten carbide. J Catal, 1998, 180(1): 85–100.
carbon source. J Solid State Chem, 2006, 179(2): 538–543.
[6] Bouchy C, Schmidt I, Anderson J R, Jacobsen C J H,
[14] Xiao T, York A P E, Williams V C, Al-megren H, Hanif A,
Derouane E G, Derouane-abd Hamid S B. Metastable fcc
Zhou X Y, Green M L H. Preparation of molybdenum carbides
a-MoC1- x supported on HZSM5: Preparation and catalytic
using butane and their catalytic performance. Chem Mater,
performance for the non-oxidative conversion of methane to
2000, 12(12): 3896–3905.
aromatic compounds. J Mol Catal A, 2000, 163(1–2): 283–296. [7] Xiang M L, Li D B, Xiao H C, Zhang J L, Li W H, Zhong B,
[15] Hianf A, Xiao T, York A P E, Sloan J, Green M L H. Study on the structure and formation mechanism of molybdenum carbides. Chem Mater, 2002, 14(3): 1009–1015.
Sun Y H. Preparation and characterization of molybdenum
[16] Zhang J Z, Long M A, Howe R F, Molybdenum ZSM-5 zeolite
carbide and its performance on the hydrogenation of carbon
catalysts for the conversion of methane to benzene. Catal
monoxide. Journal of Fuel Chemistry and Technology , 2007, 35(3): 324–328.
Today, 1998, 44(1–4): 293–300. [17] Xiao T , York A P E, Coleman K S, Claridge J B, Sloan J,
[8] Li X B, Chang Y, Wang H Y, Ma J, Wei M. Study on
Charnock J, Green M L H. Effect of carburising agent on the
n-heptane isomrization on β zeolite supported molybdenum
structure of molybdenum carbides. J Mater Chem, 2001,
carbide catalyst. Journal of Fuel Chemistry and Technology, 2007, 35(2): 228–232.
11(12): 3094–3098. [18] Mccrea K R, Logan J W, Tarbuck T L, Heiser J L, Bussell M
[9] Wang J X , Ji S F, Yang J, Zhu Q L , Li S B. Mo2 C and
E. Thiophene hydrodesulfurization over alumina-supported
Mo 2 C/Al2 O3 catalysts for NO direct decomposition. Catal
molybdenum carbide and nitride catalysts: Effect of Mo
Commun, 2005, 6(6): 389–393. [10] Edward F. Metal carbides and nitrides as potential catalysts for hydroprocessing. Appl Catal A, 2003, 240(1–2): 1–28. [11] Volpe L, Boudart M. Compounds of molybdenum and tungsten with high specific surface area: II Carbides. J Solid State Chem, 1985, 59(3): 348–356. [12] Xiao T, Wang H, Da J, Coleman K S, Green M L H. Study of the preparation and catalytic performance of molybdenum carbide catalysts prepared with C2 H2 /H2 carburizing mixture. J Catal, 2002, 211(1): 183–191. [13] Wang X H, Hao H L, Zhang M H, Li W, Tao k Y. Synthesis and characterization of molybdenum carbides using propane as
loading and phase. J Catal, 1997, 171(1): 255–267. [19] Arnoldy
P,
de
Jongej
M
C,
Moulijn
J
A.
Temperature-programmed reduction of MoO2 and MoO3 . J Phys Chem, 1985, 89(21): 4517–4526. [20] Zhu Q L, Zhao X T , Zhao Z X, Ma H J, Deng Y Q. Preparation of supported molybdenum carbide catalyst using n-hexane and its hydrodesulfurization activity. Chinese Journal of Catalysis, 2005, 26(12): 1047–1052. [21] Delporte P, Meunier F, Pham-huu C, Vennegues P, Ledoux m J,
Guille
J.
Physical characterization of molybdenum
oxycarbide catalyst; TEM, XRD and XPS. Catal Today, 1995, 23(3): 251–267.
RESEARCH PAPER
Preparation of γ-Al2O3-supported molybdenum oxycarbide catalyst and its HDS activity ZHU Quan-li1, *, ZHANG Qiu-lin2, KE Dong-xian1, YANG Xiao-bin1, CHEN Qi-feng2 1
Department of Chemistry, Hanshan Normal University, Chaozhou 521041, China
2
Chaozhou Xianglu Tungsten Industry Co., Ltd, Chaozhou 521000, China
Abstract: Investigation into the temperature programmed reduction (TPR) of bulk or γ-Al2 O3 supported molybdenum trioxide using n-hexane as a carbon source in hydrogen was carried out. T he resulting catalyst was evaluated using dibenzothiophene (DBT) as a model compound for the hydrodesulfurization (HDS). T he results show that the utilization of n-hexane can lower the temperature, to fulfill carburization to some extent. For DBT HDS, however, there exist s an optimal temperature at which carburization is carried out, leading to a high activity of the resulting catalyst. In these experiments, it has been found that the resulting catalyst exhibits the highest DBT HDS activity if the precursor, MoO3 /γ-Al2 O3, with 30% theoretic loading, is carburized at 620º C under atmospheric pressure and 0.025 mol ratio of n-hexane to hydrogen in carburizing reagents.
Key Words:
molybdenum oxycarbide; hydrodesulfurization; n-hexane; XRD
The molybdenum-based catalyst has been widely used for HDS of petroleum distillates and oil derived from coal in the industry. Thus far, its activity has been increased to a very high level via modifications in the past decades [1]. Early transition metal carbide, particularly of molybdenum and tungsten, has drawn much attention as it was reported to have catalytic activity similar to noble metal[2]. In fact, transition metal carbides have shown excellent catalytic activity in many chemical reactions [3–9]. It is even believed to be a commercial HDS catalyst of the future because of its excellent performance in HDS[10]. Molybdenum sulfide, carbide or oxycarbide catalysts can all be considered to be the result of modifications on a molybdenum-based catalyst. It is clear that different modifications, even every step of the modification, will lead to a resulting catalyst with a different surface structure, and furthermore with a different catalytic activity. Most of the carbide catalysts reported in the references have been prepared via the TPR method developed by Boudart et al[11], mostly using 20% of methane in hydrogen as the carburizing reagent. However, there are many hydrocarbons that can be used as the carbon source in carburization. The tests of C2, C3, and C4 hydrocarbons have been reported in references [12–14]. It can be inferred from these reports that the utilization of a longer chain hydrocarbon can lower the temperature, to fulfill the carburization to some extent.
Methane is the most widespread and most cheap hydrocarbon all over the world, but higher temperature is needed to transform metal oxide into carbide. C2, C3, and C4 hydrocarbons are of higher activity for carburization. However, they are put into commercial use very carefully because of the difficulty in acquiring and dealing with them. Therefore, the investigation into the possibility of liquid hydrocarbon being used as a carbon source is meaningful. In this article, the test of n-hexane in carburization of molybdenum and the activity of the resulting catalyst for DBT HDS are reported and discussed.
1 1.1
Experimental Catalysts preparation
Oxide precursors were prepared by impregnating γ-Al2O3 (20–40 mesh, BET specific area, 304 m2/g, a carrier of commercial HDS catalyst, Fushun Research Institute of Petroleum and Petrochemicals, SINOPEC) with an aqueous solution of a calculated quantity of ammonium heptamolybdate (A.R.) for 4 h. Following this step, the resulting powders were dried at 110ºC for 12 h. It was finally calcined in air at 500ºC for 3 h. Thus the oxide precursor MoO3/ γ-Al2O3 with 30% theoretical loading was obtained. Pure molybdenum trioxide was obtained by the calcination
Received: 23-Nov-2007; Revised: 28-Fan-2008 * Corresponding author. [email protected] Copyright 2008, Institute of Coal Chemistry, Chinese Academy of Sciences. Published by Elsevier Limited. All rights reserved.
ZHU Quan-li et al. / Journal of Fuel Chemistry and Technology, 2008, 36(3): 316−321
of ammonium heptamolybdate in air at 500ºC for 3 h. The apparatus used for the carburization and activity measurement was assembled by the authors. The system pressure can be controlled by a pressure regulator coupled with a back pressure regulator (TESCOM). The gas flow rate was regulated with a mass flow controller (Beijing Sevenstar Electronics Co., Ltd, P.R. China). The feeding rate of liquid materials was controlled by a liquid pump (Syltech 500). A stainless steel tubular reactor (8 mm, id) was adopted. The volatile components in the off-gas were monitored online by the Qic 20 gas analysis system (Hiden Analytical Ltd). Usually, 2.1 g of sample (ca. 3 mL) was loaded in the central part of the reactor with quartz particles (10–20 mesh) filling the spare parts. TPR of molybdenum oxide, supported or unsupported, was carried out under atmospheric pressure at a heating rate of 1ºC/min from 100ºC, to a final temperature, and then maintained at this temperature for a period of time. The ratio of n-hexane to hydrogen was achieved via changing the feeding rate of n-hexane, whereas, the hydrogen flow rate was fixed at 0.25 mol/h. In the TPR test, a Qic 20 gas analysis system was used to monitor the components in the effluent gas. Before the TPR test, the oxide precursor was dried at 115ºC for more than 12 h, and nitrogen sweeping was exempted. When the TPR was completed, the resulting catalyst was evaluated for DBT HDS directly, or it was cooled down to room temperature in hydrogen, which was followed by passivation with 1% O2 in N2 for 6 h. When DBT HDS was carried out at 290ºC under 3.0 MPa of hydrogen partial pressure, the solution of 1 mass % DBT in cyclohexane was fed at the rate of 9.4 mL/h, and the H2 flow rate was maintained at 0.12 mol/h. The liquid sample gathered at intervals of 2 h was analyzed using the ultraviolet fluorescence method (ZDS-2000, Jiangyan High-tech Analytical Instrument Co., Ltd, China) excepting the oil product gathered at the first hour. The liquid–liquid equilibrium can arrive at 97.5% or above, thus the hydrodesulfurization degree, denoted as HDS%, was simply calculated according to the following formula: HDS% =[1 – Sconcentration in sample (g-S/L)/S concentration in feedstock (g-S/L)]×100%
by X-ray diffraction using Philips X’Pert PRO diffractometer with CuKa radiation.
2 2.1
Results and discussion XRD Results
XRD patterns of molybdenum oxide precursors reduced at atmospheric pressure to different final temperatures under the condition mentioned earlier are shown in Fig. 1. It can be found in Fig. 1 that peaks at 26.0, 37.0, and 53.5o for sample a, were attributed to the characteristic diffraction peaks of MoO2 (JCPDS 86-0135). This means that MoO3 can only be reduced to MoO2. As shown in Fig. 1, the diffraction peaks at 39.5, 37.9, 34.5, 52.2, 61.6, and 69.6o for sample b, were attributed to β-Mo 2C[15], apart from the concomitant characteristic diffraction peaks of molybdenum dioxide. The XRD pattern of sample c was similar to that of sample b, but the intensity of the characteristic diffraction peaks of molybdenum dioxide was much weaker. In view of the passivation process leading to oxidation, to some extent, after carburization, it can be deemed that the carburization at 650ºC was almost complete. However, the case of the supported catalyst was different. As shown in Fig. 1 d, no characteristic diffraction peaks of molybdenum species were visible, except those of γ-alumina.
Fig. 1 XRD patterns of oxide precursors reduced at different temperatures a: ended at 550º C; b: ended at 600º C; c: ended at 650º C; d: ended at 650º C
1.2 Carburization samples preparation 2.2 Sample a, b, and c resulted from TPR of molybdenum trioxide and sample d from 30% MoO3/γ-Al2O3, using hydrogen mixed with n-hexane as the carburizing reagents at atmospheric pressure, and at a heating rate of 1 K/min from 373 K to a final temperature and held at this temperature for 2 h. When conducted, the molar ratio of n-hexane to hydrogen was kept at 0.025. The crystalline components of the catalyst were identified
TPR-MS of MoO 3 /γ-Al 2 O 3
The blank experiment was performed to detect the catalysis of the stainless steel reactor together with the quartz particles toward the decomposition of n-hexane. Fig. 2 shows that the ionic fragments resulted from volatile products tracked by the Qic 20 gas analysis system versus time on stream during temperature-programmed reduction, sweeping the reactor and quartz particles using the mixture of n-hexane
ZHU Quan-li et al. / Journal of Fuel Chemistry and Technology, 2008, 36(3): 316−321
and hydrogen. These ionic fragments shown in Fig. 2 can generally result from many molecules, but they, particularly the change in their intensity, can be limited within much lesser species if feasible reactions are taken into account. The profile of signal m/z = 78, attributed to benzene in the products, indicated that the stainless steel reactor cannot catalyze the dehydrocyclization of n-hexane to benzene below 650ºC. The profile of signal m/z = 18, attributed to water steam, indicates that some water can be removed from the quartz particles, mainly below approximately 430ºC. It is hardly formed above this temperature. Up to 600ºC, there is no change in m/z = 44 attributed to CO2 or C3 or higher hydrocarbon, which meant no reduction of quartz. A slight decrease from above 600°C can result in the decrease of C3 or higher hydrocarbon. The signal of m/z = 43 can result from C3 or higher hydrocarbon. Its decrease from above approximately 580ºC denoted n-hexane consumption rather than the substantial formation of C3 and C4 hydrocarbons; otherwise it increased with elevated temperature . The increase of m/z = 16, 28, and 30 above 600ºC, can occur in all organic compounds, which indicates the initiation of n-hexane decomposition to methane or C2 hydrocarbon, if not, there must be some heavier ionic fragments that increase with temperature. Fig. 2 shows the fact that n-hexane decomposition takes place above 600ºC in this condition, no matter how this stainless steel reactor including quartz filling catalyzes.
at its maximum at approximately 525ºC. As for the signal m/z = 44, though weak, it arrives at its highest at 550ºC. The signal of m/z = 43 shows a slight decrease in temperature before 250ºC. Following this temperature a relatively stable status is maintained up to approximately 460ºC. It decreases substantially with temperature above 460ºC.
Fig. 3 TPR – MS of MoO3 /γ -Al2 O3 The experiment was carried out under atmospheric pressure, using a carburizing reagent with 0.025 molar ratio of n-hexane to hydrogen, heating at a rate of 1ºC /min from 100ºC to 650ºC
2.3 Effect of compositions in carburizing reagents on DBT HDS activity The effect of hydrocarbon-to-hydrogen ratio in carburizing reagents on the DBT HDS activity of the resulting catalyst is shown in Fig. 4.
Fig. 2 Blank experiment The experiment was carried out at atmospheric pressure, using a carburizing reagent with 0.025 molar ratio of n-hexane to hydrogen, heating at a rate of 1ºC /min from 100ºC to 650ºC
The online analysis results of the volatile products during TPR of MoO3/γ-Al2O3 using the mixture of n-hexane and hydrogen are shown in Fig. 3. The profile of m/z=18 shows water formation with temperature. Three reduction peaks can be found at 340, 504, and 620ºC, respectively, and a should er peak at approximately 366 ºC. A peak at approximately 560 ºC occurs in the curve of m/z = 78, across the experimental temperature range. The signal m/z = 28 arrives at its highest at approximately 545ºC, whereas, the signal m/z = 30 arrives
Fig. 4 The effect of carburizing agent compositions on DBT HDS activity Catalysts that resulted from 30% MoO3 /γ -Al2 O3 carburized at atmospheric pressure using n-hexane as the carbon source, heating at a rate of 1º C /min from 100º C to 650ºC and held for 2 h at this final temperature. The feeding rate of n-hexane is shown in the legend
As mentioned earlier, the ratio of hydrocarbon to hydrogen in the carburizing reagents varied through changing the n-hexane feeding rate at a fixed hydrogen flow rate. DBT HDS activity of the resulting catalysts with time on stream in Fig. 4 showed a very high initial activity, even as high as
ZHU Quan-li et al. / Journal of Fuel Chemistry and Technology, 2008, 36(3): 316−321
97%, for every catalyst. However, these catalysts deactivated rapidly to approximately 47% after a 17-h run. However, the catalyst that resulted from carburization under the condition of 0.80 mL/h of n-hexane feeding rate and 0.025 molar ratio of n-hexane to hydrogen, guaranteed to be with higher activity within a relatively steady stage, judging from the activity trend shown in Fig. 4. For the catalyst resulting from the lowest n-hexane feeding rate, it was possible to get to a relatively steady stage of lowest activity. 2.4 Effect of carburization temperature and time on DBT HDS activity The optimal carburization temperature was about 620ºC, deduced from the results shown in Fig. 3, which will be discussed later. Therefore, several carburization temperatures scattered around 620ºC were investigated. The result is presented in Fig. 5. All catalysts showed very initial activity together with fast deactivation, as shown in Fig. 4. Judging by the activity trend, the catalyst that resulted from the carburization at 620ºC for 2 h showed a better activity than those carburized at 650ºC and 600ºC for 2 h, respectively, that is , promised to get to a steady state of a higher activity. When the final temperature ended at 620ºC, the catalyst resulted from 1 h of carburization, and exhibited a higher activity than those of 2 h and 4 h carburization, respectively.
Fig. 5 Effect of carburization temperature and time on HDS activity Catalysts that resulted from 30% MoO3 /γ -Al2 O3 carburized at atmospheric pressure using n-hexane as a carbon source with 0.025 molar ratio of n-hexane to hydrogen, heating at a rate of 1º C /min from 100º C to a different final temperature and held for different time shown in the legend at final temperature
2.5
Discussions
The pyrolysis or hydrocracking of n-hexane can lead to smaller hydrocarbon molecules, therefore, it is possible to use n-hexane as the carbon source, similar to methane, C2, C3, and C4 hydrocarbons, for the carburization of transition metal precursors . Comparing the ionic fragments resulting from
n-hexane, shown in Fig. 2 and Fig. 3, it can be found that γ-alumina supported molybdenum oxide catalyzed the n-hexane decomposition at a suitably high temperature, whether the stainless steel reactor had catalysis or not. Certainly, some transformation occurred on the surface of the molybdenum oxide with elevated temperature. In fact, the carburization of the molybdenum oxide was also a process of oxygen removal, thus the formation of an oxygen containing species could indicate the reduction state of molybdenum oxide. The profile of water formation in Fig. 3, m/z = 18, was similar to that reduced by hydrogen at a heating rate of 1 K/min in ref[16], and also similar to that carburized using butane and hydrogen [14]. Taking the XRD results shown in Fig. 1 into account, three reduction peaks below 550ºC were attributed to the process of MoO3 reduced to MoO2, whereas, oxygen removal above 550ºC was attributed to carburization. In Fig. 3, the signal of m/z = 44 could be attributed to concomitant ionic fragments of C3 or heavier hydrocarbon, as also to CO2. If it was the former, its intensity would synchronously vary with fragments such as , m/z = 43 and so on. However, the fact shown in Fig. 3 was not the case. This meant CO2 formation contributed to the increased intensity of m/z = 44. The signal of m/z = 28 could be attributed to C2 or heavier hydrocarbon, or N2 or CO. If N2 existed as an impurity, it did not change the intensity of m/z = 28. If the signals of m/z = 28 and 30 were only attributed to C2 hydrocarbon, they should arrive at the maximum at the same temperature. The results in Fig. 3 indicated that there were other species that contributed to m/z = 28 or 30. When CO2 was formed, gun with the CO formation. Prior to COx formation, the pyrolytic C or CHx group that resulted from the rupture intramolecule of n-hexane was deposited on the surface of molybdenum oxide. This led to the oxycarbide formation, namely, initiating the carburization. The signal of m/z = 78 in Fig. 3 indicated benzene formation at 560ºC, which resulted from the dehydrocyclization of n-hexane via cyclohexane intermediate over the molybdenum oxide. Accordingly, the surface of MoO3/γ-Alumina had been subjected to a process of MoO3 → MoO2, and an initiated carburization, as mentioned earlier. This dehydrocyclization of n-hexane occurred more possibly via bifunctional catalysis than by oxidative dehydrocyclization, catalyzed by a metal oxide[17], because the oxycarbide should be formed before the carbide formation[15]. Moreover, the oxycarbide rather than carbide possesses the bifunction[18]. When the oxycarbide was formed, the catalysis to hydrocracking was enhanced, and the enhanced catalysis resulted in rapid methane formation as shown in Fig. 3. The promoted deep hydrocracking in its turn enhanced the carburization. Therefore, carburization should be completed at a temperature above 560ºC. As shown in Fig. 1, the carburization was almost complete at 650ºC for bulk MoO3.
ZHU Quan-li et al. / Journal of Fuel Chemistry and Technology, 2008, 36(3): 316−321
In view of the promotion of support to the dispersion of MoO3, resulting in an easier chemical reaction, the carburization of MoO3/γ-Al2O3 would be favored. This was to say that MoO3/γ-Al2O3 could be highly carburized before 650ºC. Taking all factors into account, the temperature of 620ºC indicated by the profile of m/z = 18 signal in Fig. 3 was possibly the optimal carburization temperature for 30% MoO3/γ-Al2O3. Compared with methane, a lower temperature was needed to fulfill the carburization of the supported molybdenum oxide[19]. No matter what hydrocarbon is used as the carbon source, there is always an optimal ratio of hydrocarbon to hydrogen in the carburizing reagents for the resulting catalyst. As shown in Fig. 4, this optimal molar ratio was 0.025 for the carburization of 30% MoO3/γ-Al2O3 at 650ºC, when n-hexane was used as the carbon source, and the resulting catalyst was evaluated by DBT HDS. In view of the theoretically possible carbon deposition quantity, this value was less than 20% methane or 5% butane in hydrogen. It accounted for the higher chemical activity of n-hexane for the carburization. Of course, this value should be related to the oxide precursor, to the carburization temperature, and to the reaction the resulting catalyst should be oriented to, and so on. A higher n-hexane to hydrogen ratio in this work than reported earlier[20] was possibly attributed to the hydrogen used in the carburizing reagents. There are two parallel routes to desulfurization: one is direct desulfurization and the other is the desulfurization though hydrogenation [21]. Therefore, the surface acidity and hydrogenation capacity are needed for a good HDS catalyst. Usually, the carbide based catalyst reported in literature was a real oxycarbide catalyst because very high temperature was needed to complete the oxygen removal[19]. The difference among them was the carburization degree. The residual oxygen not only helped to stabilize the so-called carbide phase, but also helped to form B-acidic sites, to activate hydrogen or to promote the H spillover, which were helpful to HDS. The Fig. 5 indicated that the catalyst resulted from carburization at 620ºC rather than at 650ºC or at 600ºC had a better DBT HDS activity, and the carburization at 620ºC for 1 h was better than for 2 or 4 h. This result meant that some residual oxygen in the carbide catalyst was necessary to maintain a highly stable HDS activity. As reported in references [3,14] , a moderately carburized catalyst exhibited better hydrotreating activity. The higher temperature carburization resulted in the formation of a typical carbide of catalysis more similar to a noble metal, but other effects upon the activity, such as the decrease of specific area, surface acidic sites , and so on, worked at the same time. The results shown in Fig. 5 indicated that the optimal carburization temperature and time and the composition of carburizing reagents will result in an optimal synergism
among the specific area, pore distribution, surface acidity, hydrogenation ability, and so on. In fact, the molybdenum carbide catalyst in these experiments, whether carburized at 650ºC, at 600ºC or at another temperature, same as in most of the literature, was actually made up of oxycarbide. When n-hexane was used to carburize the metal oxide, a lower carburization temperature was needed because of its high chemical activity. However, it was this kind of high chemical activity that also led to substantial carbon deposition on the resulting catalyst at the same time. Thus the composition of carburizing reagents was needed to be more strictly controlled. 3
Conclusions
Primary study on molybdenum oxide carburization using n-hexane as carbon source was carried out. It was found that n-hexane could be used in the preparation of a carbide catalyst and a lower temperature was needed to fulfill the carburization. To obtain this kind of oxycarbide-based catalyst with high HDS activity, many parameters needed to be strictly controlled. For the precursor of 30% MoO3/γ-Al2O3, high DBT HDS activity was obtained if it was carburized at 620ºC for 1 h under atmospheric pressure, using carburizing reagents with 0.025 molar ratio of n-hexane to hydrogen. Certainly, the effects of many other preparation parameters on the catalytic perfomance are needed to be studied in detail.
Acknowledgment:
Zhu Quanli and his co-authors thank the
Chaozhou Xianglu Tungsten Industry Co., Ltd for the financial support rendered by it for this study.
References [1] Kabe T , Ishihara A, Qian W. Hydr odesulfurization and hydrodenitrogenation: Chemistry and Engineering. Weinheim: Wiley-VCH, 1999. [2] Levy R B, Boudart M. Platinum-like behavior of tungsten carbide in surface catalysis. Science, 1973, 181(4099): 547–549. [3] Jin G Z, Zhu J H, Fan X U, Sun G D, Gao J B. Effect of Ni promoter
on
dibenzothiophene
hydrodesulfurization
performance of molybdenum carbide catalyst . Chinese Journal of Catalysis, 2006, 27(10): 899–903. [4] Li S, Lee
J
S,
Hyeon
T,
Suslick
K
S.
Catalytic
hydrodenitrogenation of indole over molybdenum nitride and carbides with different structures. Appl Catal A, 1999, 184(1): 1–9. [5] Claridge J B, York A P E, Brungs A J, Marquez-alvarez C, Sloan J, Tsang S C, Green M L H. New catalysts for the conversion of methane to synthesis gas: Molybdenum and
ZHU Quan-li et al. / Journal of Fuel Chemistry and Technology, 2008, 36(3): 316−321 tungsten carbide. J Catal, 1998, 180(1): 85–100.
carbon source. J Solid State Chem, 2006, 179(2): 538–543.
[6] Bouchy C, Schmidt I, Anderson J R, Jacobsen C J H,
[14] Xiao T, York A P E, Williams V C, Al-megren H, Hanif A,
Derouane E G, Derouane-abd Hamid S B. Metastable fcc
Zhou X Y, Green M L H. Preparation of molybdenum carbides
a-MoC1- x supported on HZSM5: Preparation and catalytic
using butane and their catalytic performance. Chem Mater,
performance for the non-oxidative conversion of methane to
2000, 12(12): 3896–3905.
aromatic compounds. J Mol Catal A, 2000, 163(1–2): 283–296. [7] Xiang M L, Li D B, Xiao H C, Zhang J L, Li W H, Zhong B,
[15] Hianf A, Xiao T, York A P E, Sloan J, Green M L H. Study on the structure and formation mechanism of molybdenum carbides. Chem Mater, 2002, 14(3): 1009–1015.
Sun Y H. Preparation and characterization of molybdenum
[16] Zhang J Z, Long M A, Howe R F, Molybdenum ZSM-5 zeolite
carbide and its performance on the hydrogenation of carbon
catalysts for the conversion of methane to benzene. Catal
monoxide. Journal of Fuel Chemistry and Technology , 2007, 35(3): 324–328.
Today, 1998, 44(1–4): 293–300. [17] Xiao T , York A P E, Coleman K S, Claridge J B, Sloan J,
[8] Li X B, Chang Y, Wang H Y, Ma J, Wei M. Study on
Charnock J, Green M L H. Effect of carburising agent on the
n-heptane isomrization on β zeolite supported molybdenum
structure of molybdenum carbides. J Mater Chem, 2001,
carbide catalyst. Journal of Fuel Chemistry and Technology, 2007, 35(2): 228–232.
11(12): 3094–3098. [18] Mccrea K R, Logan J W, Tarbuck T L, Heiser J L, Bussell M
[9] Wang J X , Ji S F, Yang J, Zhu Q L , Li S B. Mo2 C and
E. Thiophene hydrodesulfurization over alumina-supported
Mo 2 C/Al2 O3 catalysts for NO direct decomposition. Catal
molybdenum carbide and nitride catalysts: Effect of Mo
Commun, 2005, 6(6): 389–393. [10] Edward F. Metal carbides and nitrides as potential catalysts for hydroprocessing. Appl Catal A, 2003, 240(1–2): 1–28. [11] Volpe L, Boudart M. Compounds of molybdenum and tungsten with high specific surface area: II Carbides. J Solid State Chem, 1985, 59(3): 348–356. [12] Xiao T, Wang H, Da J, Coleman K S, Green M L H. Study of the preparation and catalytic performance of molybdenum carbide catalysts prepared with C2 H2 /H2 carburizing mixture. J Catal, 2002, 211(1): 183–191. [13] Wang X H, Hao H L, Zhang M H, Li W, Tao k Y. Synthesis and characterization of molybdenum carbides using propane as
loading and phase. J Catal, 1997, 171(1): 255–267. [19] Arnoldy
P,
de
Jongej
M
C,
Moulijn
J
A.
Temperature-programmed reduction of MoO2 and MoO3 . J Phys Chem, 1985, 89(21): 4517–4526. [20] Zhu Q L, Zhao X T , Zhao Z X, Ma H J, Deng Y Q. Preparation of supported molybdenum carbide catalyst using n-hexane and its hydrodesulfurization activity. Chinese Journal of Catalysis, 2005, 26(12): 1047–1052. [21] Delporte P, Meunier F, Pham-huu C, Vennegues P, Ledoux m J,
Guille
J.
Physical characterization of molybdenum
oxycarbide catalyst; TEM, XRD and XPS. Catal Today, 1995, 23(3): 251–267.